data_kitten 1.0.1 → 1.0.2

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    ZDVlMjJmYjY5ZTE0N2I4MjVjNjNjZDQwNWZlMWFkZmNjZjUzOGIwYg==
+    OWUwYzRlMGU1MWNlYzNmZDNkZDIzYWQyOWZmZjAzNjcxZjVmMTVlZg==
   data.tar.gz: !binary |-
-    NTA2MzRkMTMzMTU4ZjA5YmE4NDRjNzZjMjYzMTU4YjdlYmM5NzZhNw==
+    M2U2M2U4Njg2MGQwYWUyMzdkMzk3YmNmMWJiMGNiNWMzODhkMTE2Nw==
 SHA512:
   metadata.gz: !binary |-
-    YTFiMWViZWUzMzVkOTVlNzM4NTNkNTIwZDY5M2FkOWI3ZDU5Mzc1MWM4NTI3
-    NTY0ZTc2ZGJmNjc2ZTEyZWE3Zjc5ODc5NTVkNmJmOWVjOTczMGJhZjQ2ZGE0
-    YTJlYWVkNzZkYjBhZjM0YTY5NTM5OTBmOWQyZjZhMzA2ODg0MjE=
+    ZmJhNmJiMmFkNDhkZjI1NmI4NDYyOGEzMDkxNjBjMzcxYTY0OGMyZWY5NWYw
+    NTUzOTI3N2FlYmU5NzQ4ZTZjN2NiNTlhMWU4NjQ3NWRhMmYwOTE3ZTAwZGE0
+    ODNiYzM4NjYzOTdmMDVmMGFjNWQ4ZTQ4ZGVkZTFiOTAyNWNlMGY=
   data.tar.gz: !binary |-
-    MjUyMTgxZDhjYWQ3NmQ4NDcwMWEzNjg3MzEwM2FmZjg4YzA1NjE2OGQ0OTYw
-    YmQwZmNiZjc0MTQ1NWJkOTE2N2NjMjgwNTQzM2VkMzljOTU5N2NjNjEwOTc2
-    YTk1YTYwOTY4ZGEyODFkZTdlZDA1NzQxZTNmNmUxZDdkOTA4NmE=
+    YzQ2YTM4YzllNDRhYzQzZjhkYWVmNjY0MThlMTZiZTg0ZjIwNWZlZDRiNmM5
+    ZmM5MWM4NDc4ZDViMDAwOWFhM2JjYTU1ZTI0YWVhNDQyYzMzZDJlYjY1MDkw
+    YzA1M2U1MmVlN2FlMmE4OGRkODAwYTA1Y2RhNTNhMGZlZDdiODk=

data/lib/data_kitten/distribution.rb

@@ -2,11 +2,11 @@ module DataKitten
 
   # A specific available form of a dataset, such as a CSV file, an API, or an RSS feed.
   #
-  # Based on {http://www.w3.org/TR/vocab-dcat/#class-distribution dcat:Distribution}, but 
+  # Based on {http://www.w3.org/TR/vocab-dcat/#class-distribution dcat:Distribution}, but
   # with useful aliases for other vocabularies.
   #
   class Distribution
-    
+
     # @!attribute format
     #   @return [DistributionFormat] the file format of the distribution.
     attr_accessor :format
@@ -34,30 +34,30 @@ module DataKitten
     #                  change to a more structured object later.
     attr_accessor :schema
 
+    # @!attribute extension
+    #   @return [String] the file extension of the distribution
+    attr_accessor :extension
+
     # Create a new Distribution. Currently only loads from Datapackage +resource+ hashes.
     #
     # @param dataset [Dataset] the {Dataset} that this is a part of.
     # @param options [Hash] A set of options with which to initialise the distribution.
-    # @option options [String] :datapackage_resource the +resource+ section of a Datapackage 
+    # @option options [String] :datapackage_resource the +resource+ section of a Datapackage
     #                                                representation to load information from.
-    def initialize(dataset, options) 
+    def initialize(dataset, options)
       # Store dataset
       @dataset = dataset
       # Parse datapackage
       if r = options[:datapackage_resource]
         # Load basics
         @description = r['description']
-        # Load HTTP Response for further use
-        if r['url']
-          @response = Curl::Easy.http_head(r['url'])
-        end
         # Work out format
         @format = begin
-          extension = r['format']
-          if extension.nil?
-            extension = r['path'].is_a?(String) ? r['path'].split('.').last.upcase : nil
+          @extension = r['format']
+          if @extension.nil?
+            @extension = r['path'].is_a?(String) ? r['path'].split('.').last.upcase : nil
           end
-          extension ? DistributionFormat.new(extension, @response) : nil
+          @extension ? DistributionFormat.new(self) : nil
         end
         # Get CSV dialect
         @dialect = r['dialect']
@@ -76,19 +76,14 @@ module DataKitten
         @title       = r[:title]
         @description = r[:title]
         @access_url  = r[:accessURL]
+        @extension   = r[:format]
         # Load HTTP Response for further use
-        if @access_url
-          @response = Curl::Easy.http_head(@access_url) do |c|
-                        c.follow_location = true
-                        c.useragent = "curb"
-                      end
-        end
-        @format = r[:format] ? DistributionFormat.new(r[:format], @response) : nil
+        @format = r[:format] ? DistributionFormat.new(self) : nil
     end
       # Set default CSV dialect
       @dialect ||= {
         "delimiter" => ","
-      }     
+      }
     end
 
     # A usable name for the distribution, unique within the {Dataset}.
@@ -112,13 +107,13 @@ module DataKitten
         end
       end
     end
-    
+
     # Whether the file that the distribution represents actually exists
     #
     # @return [Boolean] whether the HTTP response returns a success code or not
     def exists?
       if @access_url
-        @response.response_code != 404
+        http_head.response_code != 404
       end
     end
 
@@ -134,9 +129,9 @@ module DataKitten
         end
         if datafile
           case format.extension
-          when :csv 
+          when :csv
             CSV.parse(
-              datafile, 
+              datafile,
               :headers => true,
               :col_sep => @dialect["delimiter"]
             )
@@ -151,6 +146,17 @@ module DataKitten
       end
     end
 
-  end  
+    def http_head
+      if @access_url
+        @http_head ||= begin
+          Curl::Easy.http_head(@access_url) do |c|
+            c.follow_location = true
+            c.useragent = "curb"
+          end
+        end
+      end
+    end
+
+  end
 
 end

data/lib/data_kitten/distribution_format.rb

@@ -1,23 +1,22 @@
 module DataKitten
 
   # A file format for a distribution
-  # 
+  #
   # For instance CSV, XML, etc.
   #
   class DistributionFormat
-   
+
     #@!attribute extension
     #@return [Symbol] a symbol for the file extension. For instance, :csv.
     attr_reader :extension
 
     # Create a new DistributionFormat object with the relevant extension
     #
-    # @param extension [String] the file extension for the format
-    def initialize(extension, response)
+    # @param distribution [Distribution] the distribution for the format
+    def initialize(distribution)
+      @distribution = distribution
       # Store extension as a lowercase symbol
-      @extension = extension.to_s.downcase.to_sym
-      # Store response for later use
-      @response = response
+      @extension = distribution.extension.to_s.downcase.to_sym
       # Set up format lists
       @@formats ||= {
         csv:     { structured:  true, open:  true },
@@ -37,7 +36,7 @@ module DataKitten
         shp:     { structured:  true, open:  true },
         html:    { structured: false, open:  true },
         doc:     { structured: false, open:  false },
-        pdf:     { structured: false, open:  true },  
+        pdf:     { structured: false, open:  true },
       }
     end
 
@@ -54,7 +53,7 @@ module DataKitten
     def open?
       @@formats[@extension][:open] rescue nil
     end
-    
+
     # Whether the format of the file matches the extension given by the data
     #
     # @return [Boolean] whether the MIME type given in the HTTP response matches the data or not
@@ -62,12 +61,12 @@ module DataKitten
       begin
         mimes = []
         MIME::Types.type_for(@extension.to_s).each { |i| mimes << i.content_type }
-        !!(@response.content_type =~ /#{mimes.join('|')}/) || false
+        !!(@distribution.http_head.content_type =~ /#{mimes.join('|')}/) || false
       rescue
         nil
       end
     end
 
-  end  
+  end
 
 end

data/lib/data_kitten/publishing_formats/ckan.rb

@@ -171,7 +171,7 @@ module DataKitten
 
         [
           Agent.new(
-                    :name => @group["display_name"] || @group["result"]["name"],
+                    :name => @group["display_name"] || @group["result"]["title"],
                     :homepage => select_extras(@group, "website-url"),
                     :mbox => select_extras(@group, "contact-email")
                     )

data/lib/data_kitten/version.rb

@@ -1,3 +1,3 @@
 module DataKitten
-  VERSION = "1.0.1"
+  VERSION = "1.0.2"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: data_kitten
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.2
 platform: ruby
 authors:
 - James Smith
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-10 00:00:00.000000000 Z
+date: 2014-09-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake